Bayesian Logistic Gaussian Process Models for Dynamic Networks

Daniele Durante, David Dunson
Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics, PMLR 33:194-201, 2014.

Abstract

Time-varying adjacency matrices encoding the presence or absence of a relation among entities are available in many research fields. Motivated by an application to studying dynamic networks among sports teams, we propose a Bayesian nonparametric model. The proposed approach uses a logistic mapping from the probability matrix, encoding link probabilities between each team, to an embedded latent relational space. Within this latent space, we incorporate a dictionary of Gaussian process (GP) latent trajectories characterizing changes over time in each team, while allowing learning of the number of latent dimensions through a specially tailored prior for the GP covariance. The model is provably flexible and borrows strength across the network and over time. We provide simulation experiments and an application to the Italian soccer Championship.

Cite this Paper

BibTeX
@InProceedings{pmlr-v33-durante14, title = {{Bayesian Logistic Gaussian Process Models for Dynamic Networks}}, author = {Durante, Daniele and Dunson, David}, booktitle = {Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics}, pages = {194--201}, year = {2014}, editor = {Kaski, Samuel and Corander, Jukka}, volume = {33}, series = {Proceedings of Machine Learning Research}, address = {Reykjavik, Iceland}, month = {22--25 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v33/durante14.pdf}, url = {https://proceedings.mlr.press/v33/durante14.html}, abstract = {Time-varying adjacency matrices encoding the presence or absence of a relation among entities are available in many research fields. Motivated by an application to studying dynamic networks among sports teams, we propose a Bayesian nonparametric model. The proposed approach uses a logistic mapping from the probability matrix, encoding link probabilities between each team, to an embedded latent relational space. Within this latent space, we incorporate a dictionary of Gaussian process (GP) latent trajectories characterizing changes over time in each team, while allowing learning of the number of latent dimensions through a specially tailored prior for the GP covariance. The model is provably flexible and borrows strength across the network and over time. We provide simulation experiments and an application to the Italian soccer Championship.} }
Endnote
%0 Conference Paper %T Bayesian Logistic Gaussian Process Models for Dynamic Networks %A Daniele Durante %A David Dunson %B Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2014 %E Samuel Kaski %E Jukka Corander %F pmlr-v33-durante14 %I PMLR %P 194--201 %U https://proceedings.mlr.press/v33/durante14.html %V 33 %X Time-varying adjacency matrices encoding the presence or absence of a relation among entities are available in many research fields. Motivated by an application to studying dynamic networks among sports teams, we propose a Bayesian nonparametric model. The proposed approach uses a logistic mapping from the probability matrix, encoding link probabilities between each team, to an embedded latent relational space. Within this latent space, we incorporate a dictionary of Gaussian process (GP) latent trajectories characterizing changes over time in each team, while allowing learning of the number of latent dimensions through a specially tailored prior for the GP covariance. The model is provably flexible and borrows strength across the network and over time. We provide simulation experiments and an application to the Italian soccer Championship.
RIS
TY - CPAPER TI - Bayesian Logistic Gaussian Process Models for Dynamic Networks AU - Daniele Durante AU - David Dunson BT - Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics DA - 2014/04/02 ED - Samuel Kaski ED - Jukka Corander ID - pmlr-v33-durante14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 33 SP - 194 EP - 201 L1 - http://proceedings.mlr.press/v33/durante14.pdf UR - https://proceedings.mlr.press/v33/durante14.html AB - Time-varying adjacency matrices encoding the presence or absence of a relation among entities are available in many research fields. Motivated by an application to studying dynamic networks among sports teams, we propose a Bayesian nonparametric model. The proposed approach uses a logistic mapping from the probability matrix, encoding link probabilities between each team, to an embedded latent relational space. Within this latent space, we incorporate a dictionary of Gaussian process (GP) latent trajectories characterizing changes over time in each team, while allowing learning of the number of latent dimensions through a specially tailored prior for the GP covariance. The model is provably flexible and borrows strength across the network and over time. We provide simulation experiments and an application to the Italian soccer Championship. ER -
APA
Durante, D. & Dunson, D.. (2014). Bayesian Logistic Gaussian Process Models for Dynamic Networks. Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 33:194-201 Available from https://proceedings.mlr.press/v33/durante14.html.

Related Material